To improve our understanding of the mechanisms underlying alcohol abuse, experiments are proposed to test the possibility that classical conditioning, with the effects of ethanol acting as the conditioned stimulus (CS), can temporarily modify the effects of alcohol, and hence the likelihood of its abuse. The first experiment tests the speed with which such conditioning can take place when ethanol acts as the CS and fear is the conditioned response (CR) using a paradigm in which other drugs have already been shown to act as CSs. Experiment 2 then evaluates the possible role of several sources of artifact in producing the results of Exp 1. Experiments 3 and 4 test the effect of partial reinforcement and ethanol dosage on such conditioning. Experiment 5 tests the influence of variations in the CS-US interval. Experiment 6 tests whether an anxiety CR to an alcohol CS can be established when pentylenetetrazol is used as the unconditioned stimulus. The experiments are designed to elucidate the mechanisms which initiate and maintain the abuse of alcohol, and specifically the role of drugs as conditioned stimuli.